Methods for watermark decoding

Abstract

An identification code signal is hidden in a carrier signal (such as an electronic data signal or a physical medium) in a manner that permits the identification signal later to be discerned. The carrier signal can thereby be identified, or some machine responsive action can thereby be taken. The technique can be applied in video imagery embodiments to control associated video equipment, e.g. to serve as a copy control signal.

I claim:  
     
       1. In a method of decoding encoded video to extract multi-bit auxiliary data therefrom, the encoded video representing a video sequence including plural image frames, the multi-bit auxiliary data being steganographically encoded therein, the multi-bit auxiliary data thus being generally imperceptible to human viewers of the video sequence corresponding to said encoded video, said encoding taking the form of slight changes to portions of said video representing image information to thereby represent said multi-bit auxiliary data, an improvement comprising: 
       computing parameters related to the entropies of first and second signals, the first of said signals being an altered version of said encoded video; and  
       comparing said computed parameters to discern the value of at least one bit of said multi-bit auxiliary data.  
     
     
       2. The method of  claim 1  in which the second of said signals is the encoded video. 
     
     
       3. In a method of decoding encoded video to extract multi-bit auxiliary data therefrom, the encoded video representing a video sequence including plural image frames, the multi-bit auxiliary data being steganographically encoded therein, the multi-bit auxiliary data thus being generally imperceptible to human viewers of the video sequence corresponding to said encoded video, said encoding taking the form of slight changes to portions of said video representing image information to thereby represent said multi-bit auxiliary data, an improvement comprising setting a value N, and processing encoded video from N image frames to extract said multi-bit auxiliary data therefrom, wherein greater or lesser confidence in the extracted multi-bit auxiliary data can be obtained by setting N to a greater or lesser value, respectively. 
     
     
       4. The method of  claim 3  in which said decoding includes statistically analyzing the encoded video to discern the auxiliary data encoded therein. 
     
     
       5. The method of  claim 3  which further includes disabling recording capability of an associated apparatus based on said extracted auxiliary data. 
     
     
       6. The method of  claim 3  which includes reading said encoded video from an optically encoded storage disk on which it was stored in lossy compressed form, and decompressing same prior to said decoding. 
     
     
       7. The method of  claim 6  which includes reading said encoded video in MPEG-encoded form from said optically encoded storage disk, and converting said video to analog form prior to said decoding. 
     
     
       8. The method of  claim 3  in which said decoding is accomplished in the spatial image domain. 
     
     
       9. In a method of decoding an encoded video to extract multi-bit auxiliary data therefrom, the encoded video representing a video sequence including plural image frames, the multi-bit auxiliary data being steganographically encoded therein, the multi-bit auxiliary data thus being generally imperceptible to human viewers of the video sequence corresponding to said encoded video, said encoding taking the form of slight changes to portions of said video representing image information to thereby represent said multi-bit auxiliary data, an improvement comprising processing the encoded video with a pseudo-random key signal to de-randomize the multi-bit auxiliary data steganographically encoded therein. 
     
     
       10. The method of  claim 9  which includes evaluating whether copying of the encoded video is permitted, based on the decoded auxiliary data. 
     
     
       11. In a method of decoding encoded video to extract multi-bit auxiliary data therefrom, the encoded video representing a video sequence including plural image frames, the multi-bit auxiliary data being steganographically encoded therein, the multi-bit auxiliary data thus being generally imperceptible to human viewers of the video sequence corresponding to said encoded video, said encoding taking the form of slight changes to portions of said video representing image information to thereby represent said multi-bit auxiliary data, an improvement comprising, for a first frame, performing one or more dot product operations between a representation of said frame and reference data, repeating said dot product operations for each of one or more additional frames, combining the dot products thus produced, discerning at least part of the multi-bit auxiliary data from said combined dot products, and determining whether to disable recording capability of an associated video apparatus based on said discerned data. 
     
     
       12. The method of  claim 11  which further includes determining whether to disable the recording capability of the associated video apparatus based on part of said discerned auxiliary data, other of said discerned auxiliary data serving a purpose unrelated to disabling a recording capability. 
     
     
       13. The method of  claim 11  which includes reading said encoded video from an optically encoded storage disk prior to said decoding, said encoded video being stored on the disk in MPEG-encoded form, wherein the auxiliary data is discerned notwithstanding losses associated with said MPEG-encoded form.